|
Informatics Report Series
|
|
|
|
|
|
|
|
|
| Title:Incremental One-Class Learning with Bounded Computational Complexity |
| Authors:
Rowland Sillito
; Robert Fisher
|
| Date:Sep 2007 |
| Publication Title:Int. Conf. on Artificial Neural Networks (ICANN) |
| Publisher:Springer Berlin / Heidelberg |
|
Publication Type:Conference Paper
Publication Status:Published
|
|
Volume No:LNCS 4668
Page Nos:58-67
|
|
DOI: 10.1007/978-3-540-74690-4_7
ISBN/ISSN:978-3-540-74689-8
|
- Abstract:
-
An incremental one-class learning algorithm is proposed for the purpose of outlier detection. Outliers are identified by estimating - and thresholding - the probability distribution of the training data. In the early stages of training a non-parametric estimate of the training data distribution is obtained using kernel density estimation. Once the number of training examples reaches the maximum computationally feasible limit for kernel density estimation, we treat the kernel density estimate as a maximally-complex Gaussian mixture model, and keep the model complexity constant by merging a pair of components for each new kernel added. This method is shown to outperform a current state-of-the-art incremental one-class learning algorithm (Incremental SVDD [Tax2003]) on a variety of datasets, while requiring only an upper limit on model complexity to be specified.
- Copyright:
- 2008 by The University of Edinburgh. All Rights Reserved
- Links To Paper
- 1st Link
- 2nd Link
- Bibtex format
- @InProceedings{EDI-INF-RR-1274,
- author = {
Rowland Sillito
and Robert Fisher
},
- title = {Incremental One-Class Learning with Bounded Computational Complexity},
- book title = {Int. Conf. on Artificial Neural Networks (ICANN)},
- publisher = {Springer Berlin / Heidelberg},
- year = 2007,
- month = {Sep},
- volume = {LNCS 4668},
- pages = {58-67},
- doi = { 10.1007/978-3-540-74690-4_7},
- url = {http://dx.doi.org/10.1007/978-3-540-74690-4_7},
- }
|
